sap with db2 hadr and deep compression€¦ · sap with db2 hadr and deep compression – a...
TRANSCRIPT
SAP with DB2 HADR and Deep Compression –A Customer Case Study
Helmut K. C. Tessarek, DB2 Performance, IBM Toronto LabGustav Elias, Manager for Storage and Backup, OEBB
Session Number 1267
1
Click to edit Master title style
Agenda
2
Agenda
� Overview of ÖBB (Austrian Federal Railways)
� How did it start?
� SAP and HADR
� SAP and Deep Compression
� Questions
3
Click to edit Master title style
Overview of OEBB (Austrian Federal Railways)
4
Overview of ÖBB
� The biggest mobility supplier of Austria� Transportation of human beings and carriage of
goods on rail and road� Railway system of 5,690 km� > 1,200 Railway engines� > 300 Trains� > 46,000 Employees
5
Transportation NumbersBe förde rte Fa hrgä ste (in M io.)
278 280 278
433 437 443
0
50
100
150
200
250
300
350
400
450
500
2001 2002 2003 2004 2005 2006
Be förd e rte To nn e n (in M io .)
86
87
88
92
88
93
82
84
86
88
90
92
94
2001 2002 2003 2004 2005 2006
6
Information Services –Infrastructure Operations� 2 data center locations in Vienna� other servers and computer rooms across Austria� > 500 Server and Clustered Systems (AIX, Linux, Windows, VM) � > 180 TB Storage (IBM DS8300, HDS 9585, Netapp FC3020C), HP� 78 Employees
7
Click to edit Master title style
How did it start?
8
Original Environment
� 2 SAP System landscapes – SAP HR and SAP CORE� No technical problems� No massive response time problems; however, bordering on of
the SLA's� Technical requirements for SAP systems increasing constantly� Hardware S80 Cluster Systems with HACMP – end of lifecycle� Operating system AIX 4.3.3 � Database on DB2 v7.2 and Oracle 8.1.7 end of maintenance� Oracle license cost rises !!!
9
Why DB2….
� Lower license costs and maintenance costs than other competitive databases
� DB2 Administrations-Tools inclusive (eg. HADR)� Efficient database administration / configuration� Reduction of the administration tasks (reduction in support
staffing) thereby increases the system availability� Joint „Competence Centre “ in Walldorf (SAP and IBM
Partnership)� High availability = DB2 HADR
– Very quick failover times / improved Restore process – Automatic Client-Reroute– Support for Upgrades without interruption (Rolling Upgrades)– Lower network traffic - (HADR replicates only transactions!)
� DB2 HADR Proof-of-Concept
10
SAP HR Overview (Feb. 2005)
� SAP 4.6C� DB-Server S80 with 12 CPU‘s (Power3) and 12GB Memory� 4 Applications Server with 4 CPU‘s (Power3) and 6GB Memory� 8895 SAPs� AIX 4.3.3 and Oracle 8.1.7� Active-Passive Cluster using HACMP� 600 Users� 620 GB Data volume� Largest Table: 120GB (PCL2)� High Batch load on Salary Account� Other 4 Systems (Development, Test, Quality Assurance,
Training) on Power3 Basis with a total of 4620 SAPs
11
SAP CORE Overview (Feb. 2005)
� SAP 4.6C� DB-Server S80 with 6 CPU‘s (Power3) and 12GB Memory� 6 Applications Server with 4 CPU‘s (Power3) and 4GB Memory� 10875 SAPs� AIX 4.3.3 and DB2 7.2� Active-Passive Cluster using HACMP� 1500 User� 720 GB Data volume� Largest Table: 76GB (GLPCA)� High Transaction online workload� Other 3 Systems (Development, Test, Quality Assurance) on
Power3 base with a total of 3960 SAPs
12
Project Expectation
� Modern and for future protected-optimized hardware� Extended Scalability� Platform consolidation for both SAP system landscapes uniform
operating system and database system� Minimize the costs for SAP Systems
13
Click to edit Master title style
SAP and HADR
14
High Level Architecture
P570-1 P570-2
P570-3
P570-4
HACMP
HA
CM
P
DB
2 H
AD
R
Location 1
Location 2
Abo
ut 2
0 km11 LPARs
16 Proc.11 LPARs16 Proc.
SAPTSM
SAPTSM
9 LPARs8 Proc.
8 LPARs8 Proc.
15
High Availability Disaster RecoveryScenarios:
� Onsite Standby� Offsite Disaster Recovery
Approach:
� Primary & Standby systems� permanent update of the changes� Works at level of the database� Ultra-fast fail-over (no instance restart)� If the Primary Database fails, the
Standby Database will take over the Workload (HADR-Takeover)
� If the primary server becomes available, switch back over from the secondary server to primary server
� Applications become automatic re-routed (Automatic client Reroute)
� Various HADR modes of operations(SYNC, NEARSYNC, ASYNC)
Classical Active - Passive Fail-Over Cluster
DB
Ser
ver
AS
AS(CI)(DB)
Boot
Database
Boot
AS AS
CI
New HADR TechnologyFail-Over Cluster
CIAS(CI)(DB)
DB ServerStandbyD
B S
erve
rP
rimar
y
AS
Boot
AS AS
DB DB-Stby
Boot
H A D R
16
HADR Performance Tests
775HADR syncronous
685HADR nearsync
579Without HADR
881Initial Run
Runtime in secondsDescription
SGEN software component SAP_BASIS
17
Result of Migration
� instead of 19 single servers to only 2 p570 with 15 LPARs� dynamic resource allocation between the systems� Increased availability and minimized failover time by HADR� average online response time approx. 0.4 sec.� Significantly reduced diminished backup times (>12 hrs to approx. 2.5
hrs)� Significant response time and performance improvement
– Salary Account processing time reduced by half– On-line response time up to 5x faster than in the old system
– Other Batch tests 6 to 10 times improvement
� Cost reduction by savings in the license area
18
Click to edit Master title style
SAP and Deep Compression
19
18 months later…
� SAP ECC 6.0� 1.1 TB database SAP CORE ( + 50% ) � 730 GB database SAP HR ( + 65% )� average response time approx. 0.7 seconds
20
Deep Compression Procedure
� DB2 Migration DB2 8.1 to DB2 9.1� Compression of 100 largest tables� SAP Transaction COMP� over a long weekend (3 days)
21
15 largest tables (SAP CORE)
762,877,82411,990,933VIBEPP
843,024,51218,903,200COSP
753,403,52013,614,080BSAK
733,471,10412,855,941BSAS
743,498,75213,456,738SWWWIHEAD
703,519,87211,732,907COBK
743,727,61614,336,985SWPSTEPLOG
744,648,06417,877,169BSAD
744,855,04018,673,231SWWLOGHIST
856,265,34441,768,960COSS
718,866,94430,575,669ACCTIT
7217,528,44862,601,600BSIS
7118,197,24862,749,131COEP
6729,784,32090,255,515GLPCA
Spacesaving %Size in KB with CompressionSize in KB without Compression Table
22
CPU Usage in Processors
SAP CORE (Primary) Database LPAR
23
Benefits of Deep Compression
� Decrease of Storage Usage by up to 50%
�SAP CORE 1.2 TB to 625 GB�SAP HR 750GB to 490 GB
� Decrease of Backup Times by 30%� No increase of response or batch time (CPU)
24
Click to edit Master title style
Questions
?
?
?
?
?
?
?
? ?
?? ?
25
Click to edit Master title style
Backup Slides
26
MigrationPhase 1: Planning and Preparation
� Planning of the required migration times with the help of the results from the test runs (72 hours "downtime" incl. all system works, tests, approval by the User) for the production system
� Definition of the migration schedules together with the User depending on their requirements (salary account, accounting periods and accounting periods, …)
� Scheduling of the personnel resources for the migration plans on the part of IBM and ÖBB
� Construction of the project plans (A – Z: Installation of the LPAR's up to the dismantling of the old systems)
� Construction of a detailed "timetable" for the productive conversion (To-Do list of the migration steps)
27
Migration
Phase 2: Migration of SAP HR on p570
� Migration schedule: September 2005 (Week 37)� Trial rehearsal: 2 weeks before, Migration Test System� Actual Migration:
– Thursday: Migration Development System– Friday 07:00 hr: Start of Migration Production System– Saturday 15:00 hr: End of Export/Import– 22:00 hr: End of system work on SAP– Sunday 02:00 hr: End of building HADR db regarding Backup/Restore– 08:00 hr: Final work migration– 12:00 hr: Handing over to User for final tests– 16:00 hr: Release of the new system, Backup
� Wk38: Construction of quality assurance system and test system using "redirected restore" from the Backups of production system and developing system as well as migration of the training system
28
Migration
Phase 3: Migration of SAP CORE on p570
� Migration schedule: November 2005 (Week 46)� Trial rehearsal: 3 weeks before, Migration Test System� Actual Migration:
– Thursday: Migration Development system– Friday 07:00 hr: Start of Migration Production System– Saturday 18:00 hr: End of Export/Import– 22:00 hr: End of system work on SAP– Sunday 02:00 hr: End of building HADR db regarding Backup/Restore– 08:00 hr: Final work migration– 10:00 hr: Handing over to User for final tests– 18:00 hr: Release of the new system, Backup
� Wk47: Construction of quality assurance system and test system using "redirected restore" from the Backup of the production system.
29
© Copyright IBM Corporation 2008. All rights reserve d.U.S. Government Users Restricted Rights - Use, dupli cation or disclosure restricted by GSA ADP Schedule Contract with IBM Corp.
THE INFORMATION CONTAINED IN THIS PRESENTATION IS P ROVIDED FOR INFORMATIONAL PURPOSES ONLY. WHILE EFFORTS WERE MADE TO VERIFY THE COMPLETE NESS AND ACCURACY OF THE INFORMATION CONTAINED IN THIS PRESENTATION, IT IS P ROVIDED “AS IS” WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED. IN ADDITION, THIS INF ORMATION IS BASED ON IBM’S CURRENT PRODUCT PLANS AND STRATEGY, WHICH ARE SUBJECT TO CH ANGE BY IBM WITHOUT NOTICE. IBM SHALL NOT BE RESPONSIBLE FOR ANY DAMAGES ARISING OU T OF THE USE OF, OR OTHERWISE RELATED TO, THIS PRESENTATION OR ANY OTHER DOCUMENT ATION. NOTHING CONTAINED IN THIS PRESENTATION IS INTENDED TO, NOR SHALL HAVE THE EFF ECT OF, CREATING ANY WARRANTIES OR REPRESENTATIONS FROM IBM (OR ITS SUPPLIERS OR LICEN SORS), OR ALTERING THE TERMS AND CONDITIONS OF ANY AGREEMENT OR LICENSE GOVERNING TH E USE OF IBM PRODUCTS AND/OR SOFTWARE.
IBM, the IBM logo, ibm.com, AIX, DB2, DB2 Universal Database are trademarks or registered trademarks of International Business Machines Corporation in the United States, other countries, or both. If these and other IBM trademarked terms are marked on their first occurrence in this information with a trademark symbol (® or ™), these symbols indicate U.S. registered or common law trademarks owned by IBM at the time this information was published. Such trademarks may also be registered or common law trademarks in other countries. A current list of IBM trademarks is available on the Web at “Copyright and trademark information” at www.ibm.com/legal/copytrade.shtml
SAP, mySAP, SAP NetWeaver, SAP Business Information Warehouse, SAP BW, SAP BI and other SAP products and services mentioned herein are trademarks or registered trademarks of SAP AG in Germany and in several other countries.
UNIX is a registered trademark of The Open Group in the United States and other countries.LINUX is a registered trademark of Linus Torvalds.Oracle is a registered trademark of Oracle Corporation in the United States and other countries.
Other company, product, or service names may be trademarks or service marks of others.
Disclaimer